Reading and writing Research

APA DIVISION 15 CAREER ACHIEVEMENT ADDRESS

A walk through the landscape of writing: Insights from a program of writing research

Steve Grahama,b

aDivision of Educational Education & Leadership, Arizona State University; bInstitute for Learning Sciences & Teacher Education, Australian Catholic University - Brisbane, Brisbane, Australia

ABSTRACT This article is an expanded version of my presentation to Division 15 (Educational Psychology) of the American Psychological Association for the Career Achievement Award for Distinguished Psychological Contributions to Education in 2019. It provides an overview of research conducted by colleagues and I that examined the following four topics: (a) the role of writing knowledge, strategies, motivation, and skills in writing and students’ growth as writers; (b) the connections between writing, language, reading, and learning; (c) the identification of effective writing practi- ces; and (d) the current state of writing instruction in schools. For each topic, I provide examples of the logic and the different types of evidence collected in studying each area. Concluding com- ments focus on areas still in need of investigation.

It is an incredible honor to be selected as the winner of the Career Achievement Award for Distinguished Psychological Contributions to Education from Division 15 of the American Psychological Association. When I published my first paper on writing (Graham & Miller, 1979) over 40 years ago, I never envisioned receiving such an award. I had some, but not all, of the skills needed to be a successful and impact- ful scholar. My views of writing, its terrain and contours, was also limited. Through my work with my wife, Karen Harris, and a host of other scholars (see Graham, 2021), the path ahead became increasingly certain and clear over time. Although I gladly accept the honor bestow by Division 15, this award represents a collaborative effort of the research done by my collaborators, former students, and I.

The launching pad for the 40þ years of writing research described in this article stems from Karen’s and my experi- ences as special education teachers. We came into teaching with the goal of making a difference in the lives of children who found school challenging. As we started our research careers together, this focus narrowed to the domain of writ- ing, but included reading as well as learning. Over time, our interests expanded to include all developing writers, but our interest in children with special needs never diminished or disappeared.

My colleagues and I (including Karen) mainly focused our combined scholarly efforts on four interrelated and basic issues in writing. Each of these issues is examined in this article, and I share our thinking and the multiple research methods applied to studying them. The presentation of this work suggests an orderly and linear progression. Although

there is some merit to this observation, scholarly work does not typically progress from point A to B to C in a straight- forward manner, as was the case in the work described here.

My review in this article is selective rather than compre- hensive. It is not possible to present all the studies con- ducted by my colleagues and I in a single article or cover all of the different issues investigated. For example, I did not attempt to cover the research my colleagues and I undertook on writing assessment (e.g., Graham, Hebert, et al., 2016), teachers’ beliefs and knowledge about writing (e.g., Hsiang et al., 2020; Wijekumar, Beerwinkle, et al., 2019), develop- ment and testing of digital tools for writing (e.g., Boucher et al., 2017), or efforts to impact educational policy regard- ing writing (Graham, 2019). Even so, this article provides a walk through most of the writing landscape created in my program of research. This includes, in order, research on (a) factors that contribute to children’s growth as writers (including writing knowledge, strategic writing behavior, writing motivation, writing skills, and the interaction of con- textual and cognitive factors) (b) connections between writ- ing, language, reading, and learning; (c) the identification of effective writing practices; and (d) writing instruction in schools today.

Factors contributing to children’s growth as writers

Although the earliest research Karen and I pursued involved the development and testing of instructional writing practi- ces (see Graham, 1983; Harris & Graham, 1985), the starting point in this article is research I pursued with Karen and

� 2021 Division 15, American Psychological Association

CONTACT Steve Graham [email protected] Mary Lou Fulton Teachers College, Arizona State University, Benjamin Bldg. 434D, P.O. Box 871811, Tempe, AZ 85281-1811, USA. Adapted from an APA Division 15 Career Achievement Award address to Division 15 of the American Psychological Association, August, 2020.

EDUCATIONAL PSYCHOLOGIST 2022, VOL. 57, NO. 2, 55–72 https://doi.org/10.1080/00461520.2021.1951734

others on factors that facilitate students’ growth as writers. Based on Karen and my experiences as teachers, consider- ation of multiple theories of learning (e.g., Meichenbaum, 1977), and integrative reviews of the literature (e.g., Graham, 1982; Harris, 1982), we came to the conclusion that growth in writing is driven, in part, by changes in four catalysts: students’ writing knowledge, strategies, skills, and motivation. She and I spent much of our academic careers examining the validity of this proposition. The theoretical framework that guided this work was Patricia Alexander’s Model of Domain Learning (Alexander, 1998). She proposed that the road from acclimation to competence to expertise in any domain is paved by changes in strategic behaviors, acquired knowledge, and motivational beliefs. When apply- ing her model to writing, I included foundational writing skills such as handwriting, spelling, and sentence construc- tion, as part of the mix (Graham, 2006).

To study the contribution of these potential catalysts to students’ growth as writers, Karen and I established an evaluative framework involving three propositions (Graham, 2006; Graham & Harris, 2000). I illustrate these three propo- sitions with writing knowledge. If writing knowledge con- tributes to writing growth, it is reasonable to expect the following three propositions are met. Proposition one, more skilled writers are more knowledgeable about writing than less skilled writers. Proposition two, individual differences in writing knowledge predict writing performance. Proposition three, increasing writing knowledge through instruction improves the writing performance of developing writers (e.g., Graham & Harris, 2000).

For each of the four writing catalysts (e.g., writing know- ledge), I demonstrate the multiple ways my colleagues and I examined the veracity of these three propositions. As I pre- sent evidence for each of the four potential catalysts, I iden- tify which proposition was tested (proposition one, two, or three) in each study reviewed. The evidence I present is selective, but representative of the findings from the various investigations conducted. It was not possible in the space of an article to share all of the studies conducted over the last 40 years.

Writing knowledge

As they compose text, writers access many different kinds of knowledge from long-term memory (Hayes, 1996), including their knowledge of language, specialized writing knowledge (e.g., purposes and functions of different kinds of text; attributes of written words, sentences, and text organization and type; processes for planning, drafting, and producing text; intended audiences; and tools for writing), as well as knowledge about writing topics (Graham, 2018a). Writers draw on this knowledge as they conceptualize what they write, access ideas for their writing, translate these ideas into acceptable sentences, transcribe sentences into print, and reconceptualize their writing plans and text (Graham, 2018b).

In studying writing knowledge, my colleagues and I focused mostly on discourse and topic knowledge. Our

earliest study (Graham et al., 1993) assessed students’ know- ledge about the substantive and production processes used to create writing. To assess this form of discourse know- ledge, 39 fourth, fifth, seventh, and eighth grade students with learning disabilities (LD) and 29 of their typically developing classroom peers were individually interviewed. This provided a good comparison for testing proposition one (skilled writers are more knowledgeable than less skilled writers), as students with LD experience considerable diffi- culty with writing (Graham et al., 2017).

The interviewers asked students questions designed to solicit declarative (e.g., What is good writing?), procedural (e.g., How do you plan and write a paper?), and conditional writing knowledge (e.g., What special things do you need to do when writing a paper for someone three years younger than you?). As expected, typically developing students had a more sophisticated conception of writing than students with LD. They were much more likely to view writing as a sub- stantive procedure centered on the processes involved in planning, generating, evaluating, and revising text, whereas students with LD were more likely to emphasize the import- ance of form (e.g., correct spelling, neatness) when describ- ing writing. Further, older typically developing students emphasized the importance of substantive writing processes more than their younger typically developing peers. Both of these findings provided support for proposition one.

In a subsequent study, Saddler and Graham (2007) inter- viewed 10 fourth grade students whose writing was at or below the 25th percentile on a norm-referenced test and 10 fourth grade students whose writing was at or above the 50th percentile. Students wrote a story which was scored for quality and length. They were also interviewed about their knowledge of writing using many of the same questions included in Graham et al. (1993). This study provided sup- port for proposition one, as the description of writing gener- ated by stronger writers included more examples of substantive writing processes than the descriptions from less skilled writers. The study also provided support for propos- ition two (individual differences in writing knowledge pre- dicts writing performance). For students identified as stronger writers (50th percentile or above), there were posi- tive associations between the number of substantive writing processes identified during the interview and the length and quality of their stories (correlations of 0.79 and 0.73, respectively). Support for this second proposition, however, must be tempered by the finding that associations between writing knowledge and performance of less skilled writers (25th percentile or below) were small and negative (–0.01 to �0.22).

Unfortunately, the Saddler and Graham (2007) study had two important limitations which were rectified in subse- quent investigations. It included a small and discontinuous sample, and there was no control for other variables which might mediate the relation between writing knowledge and performance. By comparison, 50 fifth grade students were interviewed about their writing discourse knowledge in a study by Olinghouse et al. (2015). This investigation used interview questions from the earlier Graham et al. (1993),

56 S. GRAHAM

and participating students also completed a multiple-choice measure assessing their knowledge about outer space. Students were asked to write three papers about this topic: a story, a persuasive paper, and an informational text. The compositions were scored for overall quality and number of genre elements. Students were selected so they represented the range and distribution of writing skills typically found in their teachers’ classes. We collected data on five control vari- ables which were likely to mediate the association between writing knowledge and writing performance: gender, interest in the topic of outer space, length of each paper, handwrit- ing fluency, and spelling. Students’ interviews were scored for the following types of discourse knowledge: substantive and production processes, knowledge of how text are organ- ized, and knowledge of genre elements in stories, persuasive, and informational writing.

This study extended Saddler and Graham (2007) by dem- onstrating that students’ discourse and topic knowledge pre- dict writing performance, providing additional support for proposition two. After controlling for variance due to the control and topic knowledge measure, discourse knowledge accounted for 12.9% to 23.3% of the variance in the quality of the three types of text and 13.8% to 19.4% in the number of genre elements included in these three papers. Additionally, topic knowledge accounted for 5.9% to 14.2% of the variance in the quality of the three texts after accounting for variance due to the control and discourse knowledge measures, although it only accounted for unique variance (38.2%) in the inclusion of genre elements in informative writing. Collectively, this study, Saddler and Graham, and other investigations (Graham et al., 2019; Olinghouse & Graham, 2009; Wijekumar, Beerwinkle, et al., 2019) provided convincing evidence that writing knowledge predicts writing performance.

I have not conducted empirical intervention studies test- ing proposition three (increasing writing knowledge through instruction improves writing). My colleagues and I have, however, conducted meta-analyses where the average- weighted effect sizes (ES) of such studies were determined. In Graham et al. (2012), teaching students the purpose and structure of different kinds of text increased the quality of writing of students in grades two to six by more than one- half of a standard deviation in nine investigations (ES ¼ 0.59). In a separate meta-analysis (Graham, Liu, Bartlett, et al., 2018), the quality of elementary and secondary stu- dents’ writing improved in eight studies (ES ¼ 0.67) as a result of a second knowledge inducing strategy where youngsters observed other youngsters’ interact with text (e.g., using it or interpreting it). In an earlier meta-analysis by Graham and Perin (2007), the quality of writing pro- duced by students in grade four to 12 in six experiments was enhanced (ES ¼ 0.25) through a third knowledge induc- ing strategy where students were directed to analyze and emulate model texts.

The research and scholarly analyses reviewed here pro- vide support for the critical role of writing knowledge in students’ writing. Although all three of the propositions examined were supported by the scholarly efforts of my

colleagues and I, additional research is needed to explore more fully how and when writing knowledge paves the road to writing growth. This includes creating a more robust evi- dence base for all three propositions, and expanding this research across all grades and different student groups.

Strategic writing behavior

There is no single approach to writing, and a single author can approach the same writing task in multiple ways (Plimpton, 1989). Even so, highly skilled writers use many of the same strategic behaviors when they compose. This includes goal setting, planning, information seeking, organiz- ing, transforming, monitoring, reviewing, evaluating, and revising to name some of the strategies they apply (Zimmerman & Risemberg, 1997).

The studies my colleagues and I conducted addressing proposition one (skilled writers are more strategic than less skilled writers) only examined one side of this equation: What are the strategic writing behaviors of students who find writing challenging? This research mostly involved elementary grade students with LD who also experienced writing difficulties. In our research, these students’ approach to writing was radically different from the approach attrib- uted to more skilled writers who plan what to say and how to say it, translate these plans into written text, and revise their plans and text to ensure they meet their writing goals (Hayes, 1996). This was illustrated in a study conducted with 12 fourth- and 12 sixth-grade students with LD (Graham, 1990). When composing opinion essays, they con- verted the writing task into a question to be answered, beginning their papers with a “yes” or “no” and quickly writing whatever came to mind, and abruptly ending the task without a concluding statement. They did not plan in advance and spent only about six minutes writing. They had access to more ideas than the ones appearing in their text, though, as prompts to “say more” doubled to tripled the amount of relevant text produced. In essence, their approach to writing involved a single writing process, generating con- tent, but this approach was generally unproductive as stu- dents concluded the writing process prematurely before sharing in writing all they knew.

In other studies, my colleagues and I demonstrated that how students go about the process of composing predicted how well they write (proposition two). A study by Graham et al. (2019) with 286 fifth grade students illustrated the pre- dictive value of students’ strategic behaviors. Students were asked to write an opinion article after reading informational text about the topic. Their compositions were scored for writing quality and length. The quality of students’ plans for writing were also assessed, as was their approach to writing using a self-report measure examining 10 aspects of writing behavior (e.g., I keep my topic and theme clearly in mind as I write). After controlling for variance due to gender, read- ing skills, self-efficacy, attitude toward writing, writing knowledge, transcription skills (e.g., spelling), and clustering at the classroom level, the two strategic writing measures (planning and approach to writing) uniquely predicted

EDUCATIONAL PSYCHOLOGIST 57

writing quality and length of composition at two time points (Fall and Spring), accounting for 1% to 9% of the variance in writing scores. Planning scores were particularly predict- ive of students’ writing performance. This measure signifi- cantly and uniquely predicted writing quality at both time points. Approach to writing did not make such a singular contribution. Studies by Wijekumar, Graham, et al. (2019) with 175 fifth-graders and Graham et al. (2017) with 227 fourth-graders reported similar effects, with planning and approach to writing accounting for up to 16% of the vari- ance in students’ writing scores.

The aspect of students’ strategic writing behavior my col- leagues and I have most thoroughly investigated involves proposition three (teaching students to be more strategic writers improves their writing). Over the years, Karen and I, our students, as well as other colleagues investigated the value of teaching a variety of different types of students in different grades strategies for planning, revising, and self- regulation (e.g., self-monitoring). In virtually all studies con- ducted to date, instructional efforts to enhance strategic writing behaviors resulted in improved writing performance. I provide three examples to demonstrate these efforts.

One, in a single-participant design study with six third- to fifth-grade students with ADHD, Harris et al. (2005) found that teaching these children to monitor how often they practiced learning spelling words or how often they were on-task when practicing these words enhanced learning of new spelling words. Two, Page-Voth and Graham (1999) demonstrated the persuasive writing of seventh- and eighth- grade students with LD was improved by teaching them a strategy to help them refute counterarguments in their paper (n¼ 30). Three, in a study with 262 second and third grade children, Harris et al. (2012) improved the quality, length, and completeness of students’ stories and persuasive essays by teaching them genre- specific strategies for planning, drafting, and self-regulation.

Most of the strategic writing intervention studies con- ducted to date focused on teaching students task-specific writing strategies and procedures for self-regulating the tar- geted writing strategies, the writing process, and writing behaviors. This approach, which Karen Harris developed and named Self-Regulated Strategy Development (SRSD; Harris & Graham, 2009) has been applied in over 100 stud- ies to date in countries all around the world with students of all types and ages and a variety of different kinds of writ- ing (see Harris et al., 2012 for an example). It is likely the most experimentally investigated writing intervention in the literature.

In the most recent meta-analysis of SRSD studies (Graham et al., 2013), I examined 84 true-experiments, quasi-experiments, or single participant design studies test- ing the effectiveness of teaching students to be more stra- tegic via SRSD. These studies involved typically developing writers, students who experience difficulty learning to write, and students receiving special education services. For true- and quasi-experiments, SRSD was effective, as ESs of 1.75, 2.24, and 0.47 for writing quality, genre elements, and length, respectively. SRSD was highly effective across

different types of students, genres, and grades in this meta- analysis. Single participant design studies evidenced similar effects for the same measures.

The evidence my colleagues and I collected provided strong support for the importance of strategic writing behaviors to students’ growth as writers. Much of this research has concentrated on planning and to a lesser degree on revising and specific aspects of self-regulation. There is a need for research on other strategic behaviors such as goals setting, self-evaluation, self-instructions, and self- reinforcement.

Writing motivation

Motivational beliefs about writing theoretically influence writers’ engagement, effort, and actions (Graham, 2018a; Hayes, 1996). These beliefs can determine whether writers are generally more or less motivated to write, or whether they are more or less motivated about a particular writing task or topic. Motivations for writing include beliefs about why one engages in writing, the importance of trying to out- perform others, the value and utility of writing, attitudes and interest in writing, competence as a writer, determinants of writing success or challenges, writing identity or identi- ties, and views about the communities where writing takes place (Graham, 2018b).

My research with colleagues focused on three specific types of motivational beliefs about writing: self-efficacy, atti- tude toward writing, and motives for writing (intrinsic, extrinsic, and self-regulatory). Our findings about the role of writing motivation as a catalysts for promoting writing growth are less consistent and more contradictory than those for writing knowledge and strategies. For example, when comparing older and younger writers as well as stu- dents who do and do not experience difficulty learning to write (proposition one), findings are mixed. To illustrate, Graham et al. (1993) found older seventh and eighth grade students (n¼ 47) had less positive attitudes about writing than younger fourth-grade students (n¼ 21), and students with LD (n¼ 39) had less positive attitudes toward writing than their typically developing peers (n¼ 29). However, there was no difference between these student groups in terms of their efficacy for writing.

These discrepancies were evident in other studies as well. Olinghouse and Graham (2009) reported fourth grade stu- dents (n¼ 32) had less positive attitudes toward writing than second grade students (n¼ 32). In a recent study com- paring emergent bilingual sixth to eighth grade students (n¼ 94) with their native English speaking classmates (n¼ 285) who scored higher on a District writing test, emer- gent bilingual students were more motivated to write for intrinsic, extrinsic, and self-regulatory reasons than their native speaking peers (Camping et al., 2020). The only motive for writing where native English speakers were more motivated was writing to obtain better grades.

Although my colleagues and I have found individual dif- ferences in motivational beliefs predict writing performance (proposition two), this was not consistently the case. For

58 S. GRAHAM

example, in the Camping et al. (2020) study, motives for writing predicted scores on the District writing test for native English speakers (accounting for 6% of the variance), but not for emergent bilingual students. In a large study with fourth and fifth graders (n¼ 2,186), Graham et al. (2021) found intrinsic, extrinsic, and self-regulatory motives for writing were associated with students’ performance on an opinion writing for a District writing test, but only extrinsic motives predicted students’ performance on a simi- lar State assessment. In both of these investigations, how- ever, motives to write accounted for just 2.1% and 2.6% of the variance in writing scores after controlling for gender, grade, race, SES, and special education status.

Somewhat similarly, multiple studies (e.g., Graham et al., 2017, 2019; Wejikumar et al., 2019b) reported attitude toward writing and self-efficacy collectively accounted for unique variance in the quality and length of persuasive text and per- sonal narratives. Even so, the amount of variance accounted for by these two motivational beliefs after controlling for stra- tegic behaviors, gender, and other variables (e.g., reading, writing knowledge, and writing skills) was quite limited (3% or less). Likewise, Limpo et al. (2020) obtained small and mostly negative associations between measures of self-efficacy and motives for writing and writing quality, length of papers, and handwriting fluency for 193 third graders.

In contrast, other studies my colleagues and I conducted provided evidence that motivational beliefs were robust pre- dictors of writing performance. In a study with 321 sixth graders, Rocha et al. (2019) reported attitude toward writing, self-efficacy, and intrinsic, extrinsic, and self-regulatory motives for writing collectively accounted for 35% of the variance in writing quality scores. Graham, Daley, et al., (2018) found attitudes toward writing and writing self- efficacy accounted for 7.3% of the variance in scores of 185 middle school students on a norm-referenced writing test after controlling for variance due to gender, SES, language status (native speaker or emergent bilingual), disability sta- tus, and self-efficacy and attitude toward reading. In this study, writing beliefs accounted for even more variance in the writing scores of students with disabilities (16.3%) than they did for typically developing students (10.0%).

There are surprisingly few studies assessing proposition three (enhancing motivation through instruction improves writing). In a meta-analysis of writing intervention research with school-aged students, Graham, Harris, et al. (2016) located three studies where researcher put into place proce- dures to enhance students’ motivation (e.g., procedures for increasing self-efficacy) and also measured improvements in students’ writing. The impact of these studies on the quality of students’ writing was substantial (ES ¼ 1.07). Despite this large effect, I suspect future studies in this area will yield more variable results, just as other studies reviewed looking at propositions one and two above did.

I base this prediction on studies where my colleagues and I implemented procedures (e.g., goal setting) which should enhance one or more aspects of motivation (e.g., self-efficacy). The effects on students’ writing beliefs in these investigations were inconsistent (see Graham & Harris,

1989; Page-Voth & Graham, 1999; Sawyer et al., 1992). In other studies, the outcomes were limited, as the effects of interventions to enhance motivation were restricted to spe- cific motivational beliefs. For instance, in a study with 206 fifth and sixth grade students (De Smedt et al., 2019), we anticipated that students who wrote with a partner over a five-week period versus students who wrote alone would be more self-efficacious and autonomously motivated. Writing with a peer gave these students multiple opportunities to observe another writer, and provided them with information for judging their own competence. We reasoned this inter- vention should increase students’ need for relatedness through working with a peer, which in turn should promote autonomous motivation. Wheras peer writing did foster greater autonomous motivation, it only did so when stu- dents in this condition were compared to children in a busi- ness-as-usual condition and not to students who received the same experimental instruction but wrote alone. Further, writing with a peer did not enhance students’ self-efficacy. The mixed effects illustrated here suggest that positively changing students’ writing motivations is an uncertain and poorly understood process, and additional research is needed to develop a more nuanced understanding of how motivational beliefs might foster students’ growth as writers.

These findings provide tentative support for the conten- tion that writing motivational beliefs are important ingre- dients in students’ growth as writers. However, this thesis is far from certain because some motivational beliefs decline with experience, motivational beliefs are not especially robust predictors of writing performance in some studies when other mediating variables are first controlled, and there is only a small number of investigations examining if efforts to improve motivation enhance writing performance too. Clearly, motivation and writing is an area of research in need of considerably more attention. This includes examin- ing a broader array of writing motivational beliefs with stu- dents at all ages and with varying characteristics, with a particular emphasis on identifying effective means for simul- taneously enhancing writing motivation and performance.

Writing skills

Adult writers typically take for granted foundational writing skills like handwriting, spelling, and typing. These transcription skills, however, can interfere with writing and may even impede children’s growth as writers (Graham & Harris, 2000). Until students can spell, type, and write letters fluently and accurately, transcribing ideas into written text can disrupt and interfere with writing in multiple ways. Having to think about how to spell a word may cause a writer to lose ideas or plans held in working memory. If handwriting and typing are not fast enough to keep up with a writer’s thinking, ideas and plans can slip away before they are recorded. Having to apply conscious effort and attention to any of these skills means there are fewer resources available for other writing actions. Students who experience difficulty mastering these skills may avoid writing whenever possible, leading to less growth as a writer over time (Berninger, 1999).

EDUCATIONAL PSYCHOLOGIST 59

Text transcription skills are not the only writing skills which are important to writing growth. For instance, writers spend a considerable amount of effort and resources on translating their ideas into acceptable sentences (Graham, 2018a). With the exception of a study by Saddler and Graham (2007) that demonstrated sentence combining instruction improved the sentence construction skills of more and less skilled fourth grade writers (n¼ 44) and the quality of their story writing too, my research with col- leagues has concentrated almost exclusively on two specific transcription skills: handwriting and spelling.

In studies with typically developing students, my col- leagues and I gathered evidence which supports proposition one (more skilled writers are faster handwriters and better spellers than less skilled writers). Graham et al. (1998) reported the handwriting fluency of students in grades one to nine (n¼ 900) increased by about 10 letters per minute from one grade to the next, leveling off at the start of high school (Graham et al., 1998). Similarly, Skar, Lei, & Graham (in press) indicated the handwriting fluency of students (n¼ 4,950) almost doubled between grade one and three. Further, in several meta-analyses, we found students who are typically weaker writers (children with LD, ADHD, speech and hearing difficulties, and reading difficulties) scored lower on measures of spelling and handwriting than their typically developing peers (Graham, Aitken, et al., 2020; Graham, Fishman, et al., 2016; Graham et al., 2018, 2020). The average difference between these students and their same age peers ranged from 0.80 to 1.50 standard devi- ations for spelling and 0.62 to 0.78 standard deviation for handwriting.

My research with colleagues also provides evidence which supports proposition two (individual differences in handwrit- ing and spelling predict writing performance). In a study with 600 grade one to six students, Graham et al. (1997) reported handwriting fluency and spelling accounted for 25% and 66% of the variability in writing quality and out- put, respectively, in grades one to three. For older students (grades four to six), these same measures accounted for 41% and 42% of the variance in writing quality and spelling, respectively. Whereas handwriting fluency contributed dir- ectly to the prediction of writing performance in the pri- mary and intermediate grades, spelling only made a direct contribution in grades one to three. Its contribution to pre- dicting writing performance in the intermediate grades was indirect through its association with handwriting.

In a more recent study, handwriting fluency again pre- dicted the quality of writing produced by 4,950 students in grades one to three (Skar, Lei, & Graham in press). The amount of variance accounted for was smaller though (7.4%), probably because we controlled for variance due to attitude toward writing, language background (L1, L2, bilin- gual), gender, grade, and nesting at the class- and school- level. Other studies with students in grade two to five (Graham et al., 2019; Olinghouse et al., 2015; Olinghouse & Graham, 2009) found handwriting fluency uniquely pre- dicted writing performance after controlling for variables

ranging from gender, reading skills, spelling, writing know- ledge, and writing motivational beliefs.

In contrast, spelling has been a less reliable predictor of writing performance in studies where variance due to these same types of variables were first controlled. For example, in Wijekumar, Graham, et al. (2019), the spelling of 175 fifth grade students uniquely accounted for 3% of the variance in persuasive length in the Fall and 2% of the variance in writ- ing quality in the Spring after controlling for gender, read- ing, writing motivation, and writing knowledge, Nevertheless, spelling was not significantly related to writing quality in the Fall and writing length in the Spring. When controlling for the same variables in the Spring plus hand- writing fluency in the Fall, Graham et al. (2019) found the spelling of fifth grade students (n¼ 229) uniquely predicted persuasive writing quality at both time points, but it was not uniquely associated with length of persuasive essays at either time points. In a study by Olinghouse et al. with fifth grade students (n¼ 50), spelling uniquely predicted the quality of persuasive text, but it did not predict the quality of stories or informational text after controlling for gender, writing output, handwriting fluency, writing motivation, and writing knowledge. Finally, Olinghouse and Graham (2009) reported that spelling did not uniquely predict story quality, length, or vocabulary of grade two and four students (n¼ 64) after variance due to grade, gender, reading, handwriting fluency, writing motivation, and writing strategies was controlled.

To test the veracity of proposition three (teaching hand- writing and spelling enhances other aspects of writing), my colleagues and I conducted studies with primary grade stu- dents where handwriting and spelling were taught together or alone to determine if such instruction enhanced the tar- geted transcription skills and other aspects of writing. These studies produced mixed results. For example, in a study with 30 first grade students experiencing difficulty mastering transcription skills, combined handwriting and spelling instruction improved students’ handwriting, spelling, sen- tence construction skills and diversity of vocabulary in their compositions, but not the length or quality of their compo- sitions (Graham, Harris, & Adkins, 2018). When just hand- writing skills were taught (n¼ 38), first grade students experiencing difficulties with handwriting evidenced improvements in this skill, sentence construction, story length, but not story quality (Graham et al., 2000).

Berninger et al. (1997) also found handwriting instruction improved the handwriting and sentence construction skills of first graders experiencing difficulties mastering this skill (n¼ 144). When only spelling skills were taught (n¼ 60), the spelling of second grade students with spelling difficul- ties improved, as did their sentence construction skills, but the intervention had no effect on story quality or length (Graham et al., 2002). Lastly, Berninger et al. (1998) found only one of their spelling interventions (phoneme letter associations taught) enhanced growth in spelling and length of compositions produced by second grade students with difficulty mastering spelling (n¼ 128).

The picture of the effects of handwriting and spelling instruction on students’ writing shifts somewhat when

60 S. GRAHAM

findings from two meta-analyses are considered. In Santangelo and Graham (2016) seven studies were located with students in grades one to nine where handwriting was taught and some aspect of writing measured. The ESs for this analysis supported proposition three, as writing quality, length, and fluency improved by 0.84, 1.33, and 0.48 stand- ard deviations. In contrast, a meta-analysis by Graham and Santangelo (2014) did not find that spelling instruction sig- nificantly improved the writing of students in grades one to six in six studies (ES ¼ 0.19). Collectively, the findings from the studies reviewed above and meta-analyses testing the three propositions provide strong support for the contention that handwriting fluency is an important contributor to young students’ growth as writers. The role of spelling is less certain given the uneven relationship between this skill and writing performance and the relative weak impact of spelling instruction on students’ writing. Additional research, especially experimental treatment studies, are needed to explore the relations between handwriting, spell- ing, sentence construction, and other foundational writing skills and students’ writing growth.

More than writing knowledge, strategies, motivation, and skill

A few years ago, I proposed a model of writing based on the following proposition: writing is simultaneously and interactively shaped and constrained by the communities where it takes place and the cognitive capabilities and resources of those that produce and teach it (Graham, 2018a, 2018b).1 Accordingly, writing and the acquisition of writing depends on how the communities in which writers participate actualize their purposes, values, norms, audien- ces, tools and actions, social and physical arrangements, motivations, power and responsibilities, stance/identity, and collective history. These are in turn influenced by historical, political, cultural, social, and institutional forces which oper- ate outside of these communities. Those who produce writ- ing in these communities draw on a variety of cognitive resources. These cognitive resources encompasses produc- tion processes for creating text, knowledge and beliefs held in long-term memory, as well as control mechanisms which enable writers to direct and regulate the writing enterprise. These control processes include attention, working memory, executive control, and self-regulation procedures. The use of these cognitive resources is moderated by a writer’s emo- tions, personality traits, and physiological states as well as the characteristics, capacity, and variability of the commu- nity where writing takes place and is learned.

As this writer(s)-within-community (WWC) model sug- gests, writing and students’ growth as writers depends on much more than just the writing knowledges, strategies, motivations, and beliefs individual writers develop and pos- sess. This has led me and my colleagues to investigate the

role of other contributors to children’s growth as writers. I share two examples here in the hope that others will con- tinue and expand this work.

Skar, Graham, & Huebner (in press) conducted an experiment to examine the effects of Covid-19 and the sub- sequent six week lockdown in Norway on first grade stu- dents’ writing. We administered measures of writing quality, handwriting fluency, and attitude toward writing to 1,636 first grade students in May and June of 2019 before the pan- demic began. In May and June of 2020, we administered the same measures to 817 new first-graders students attending the same schools. During the intervening year Covid-19 became a National Health crises and schools were closed for six weeks, with all instruction delivered online. After con- trolling for variance due to national test scores, school size, proportion of certified teachers, students per special educa- tion teacher, school hours per student, student gender and native language, we found students attending first grade during the pandemic and school lockdown had lower writ- ing quality, handwriting fluency, and attitude toward writing scores than first grade students tested a year earlier before COVID-19 emerged. This study provides an illustration of how historical factors and the resulting political decisions can influence students’ growth as writers.

My second example involves the study of teacher practi- ces. The WWC model (Graham, 2018a) proposed the writ- ing practices of teachers differ from one classroom to the next, especially if there are differences in institutional, social, political, cultural, and historical factors. Because teachers are so integral to students’ growth as writers, it is important to determine how much variability exist from one locale to the next. To assess this, my colleagues and I conducted a study with 1,313 seventh to ninth grade teachers randomly chosen from all schools in Shanghai, Hong Kong, and Taipei city. Although teachers in these different locations shared a simi- lar Chinese heritage, the participating cities evidenced differ- ent social, institutional, political, and historical backgrounds which should create variability in their teaching practices. This proved to be the case, as teachers in one locale differed from teachers in another locale in the assignment of writing, what they taught, and how they supported students’ writing. These differences were mostly a matter of degree though and not a reflection of major differences in teachers’ writing programs. Thus, consistent with the WWC model variability was evident in teachers’ writing practices, but this variability was constrained. Additional research is needed to explore why differences were not more pronounced. For instance, this may be a consequence of the globalization of writing methods and/or common views in Chinese society about writing, its importance, and teaching it. Other studies involving teachers and contextual differences are included in Graham (in press-a).

Connections between writing, language, reading, and learning

In addition to examining the role of writing knowledge, strategies, motivation, skills and contextual factors in

1The idea that students’ performance is a consequence of the interaction between context and cognitive capabilities is not new, and is represented in more general aptitude-within-situation models proposed by Snow (1994) or the thinking of theorist such as Anna Sfard (2016).

EDUCATIONAL PSYCHOLOGIST 61

students’ growth as writers, my colleagues and I also studied how writing is connected to language, reading, and learning. For example, it is commonly assumed writers draw on their language capabilities as they write (Shanahan, 2006; Sperling, 1996). Accordingly, strong language skills provide a firm foundation for writing, but difficulties with language may manifest themselves in writing in adverse ways. This led me to examine if children with speech and language dif- ficulties experience difficulties with writing which exceed those of their same-aged and language-matched peers.

My interests in the connections between writing and reading as well as learning stem in part from the hypothe- sized facilitative effect of writing and writing instruction on reading (Shanahan, 2006) and the use of writing as a tool for enhancing students’ learning of content material (Bangert-Drowns et al., 2004). As establish later in this art- icle, writing plays a relatively minor role in the daily oper- ation of most classrooms, and evidence that writing promotes reading and learning has the potential to change this equation. Further, reading and reading instruction pro- vide a potential means for enhancing students’ writing (Graham, 2020).

My colleagues and I have conducted empirical studies examining specific aspects of these relations. This includes the impact of writing on content learning (Gillespie et al., 2017; Hebert et al., 2014), shared knowledge between read- ing and writing (Proctor et al., 2020), and teaching students how to read source material when writing persuasive text (Harris et al., 2019). The predominance of my work in writ- ing, language, reading, and learning connections mostly involves meta-analyses however. I share findings from these reviews next.

Language and writing

There are good reasons for anticipating that students who experience speech and language impairments (SLI) also experience difficulties with writing (Shanahan, 2006). First, both speaking and writing rely on many of the same under- lying cognitive resources (e.g., phonological, morphological, sematic, syntactic, and pragmatic knowledge) and apply similar production processes (e.g., generating and monitor- ing intentions and production, translating messages/ideas into acceptable language, bringing this language to life by articulating it or putting it into written form). As a result, difficulties with language acquisition may manifest them- selves in children’s writing in multiple ways. Second, speak- ers and writers engage in conversations with their audiences, blurring the distinction between these two skills as each is inherently conversational (Bereiter & Scardamalia, 1982). Success in speaking with others can provide writers with potential insights which can be used when writing. Such successes may be less frequent for children with SLI, reduc- ing the potential impact of speaking on writing.

To determine if preschool to grade 12 students with SLI also experienced writing difficulties, my colleagues and I conducted a meta-analysis comparing the writing of students with SLI to typically developing same-aged peers and peers-

matched on language capabilities (Graham et al., 2020). I located 39 studies making one or both of these comparisons. When all writing measures were included in the analysis col- lectively, students with SLI scored significantly lower (effect size ¼ �0.97) than their same aged-peers (39 studies). More specifically, they scored lower on measures of writing quality (–0.92), writing output (–1.00), spelling (–1.14), grammar (–0.68), and vocabulary (–0.68). When students with SLI were compared to younger language-matched peers (six studies), they continued to score lower on all writing meas- ures collectively (–0.43). Because of the small number of studies, we did not examine effect sizes for specific writing measures for SLI vs language-matched peers.

These findings provide support for the contention that writing is language-based and difficulties with specific aspects of language are reflected in children’s writing. In effect, language disorders are inherently a learning disorder which also impacts students’ writing. Surprisingly little research has focused on identifying effective writing practi- ces for students experiencing difficulties with language acquisition (Graham et al., 2020), and little is known about whether teaching language skills results in better writing. There are a small number of studies showing vocabulary instruction improves writing (see Graham, Harris, et al., 2016), and even more studies showing phonological aware- ness instruction enhances spelling (Graham, Liu, Bartlett, et al., 2018). Studies examining the effects of language instruction on writing need to become more common.

Writing and reading

“Reading and writing are connected at the most intimate level. We write so others will read, and read what others write” (Graham, 2020, p. 37). People read what others write to acquire information for their own text, and they reread what is written to make sure the text created captures their intentions and is as error free as possible. Theoretically, writing and reading are connected in at least three other ways (Shanahan, 2006). One, writers and readers draw on overlapping bodies of knowledge to carry out processes such as decoding and encoding as well as meaning making and meaning comprehension. Two, reading and writing involve conversations between those who produce and consume the printed word. This provides opportunities for new insights, as when readers consider why an author used certain word, phrase, sentence, or rhetorical device. In essence, reading presents students with models for their own writing. Three, writing and reading can be used together in functional ways to meet specific goals such as learning.

To date, my colleagues and I have conducted five meta- analyses providing support for the following theoretical con- tentions: (a) reading and writing are connected, (b) and teaching and engaging in either one of these skills enhances the other. In our most recent review (Graham, Aitken, et al., 2020), 87 studies were located comparing preschool to grade 12 students experiencing difficulties learning to read with typically developing peers matched by age, and 24 investiga- tions were found comparing them to younger reading-

62 S. GRAHAM

matched peers. Children with established difficulties with reading had writing scores 1.25 and 0.94 standard deviations lower than their same-aged peers and younger reading- matched peers, respectively, when all writing measures were included in the analyses. For the first comparison with same-aged peers, spelling differences were especially pro- nounced (ES ¼ �1.42), followed by written vocabulary (ES ¼ �1.17) syntax (–1.07), writing quality (ES ¼ �0.95), sen- tence skills (ES ¼ �0.78), organization of content (ES ¼ �0.72), writing output (ES ¼ �0.66), and handwriting (ES ¼ �0.64). For the second comparison with reading-matched peers, students experiencing difficulties with reading scored �0.94 standard deviations lower on measures of spelling. The findings from this meta-analysis provide evidence sup- porting the contention that stronger readers are better writ- ers than weaker readers.

Other meta-analyses examining reading and writing con- nections provided support for the proposition that instruc- tion to improve reading improves writing and vice versa. An initial starting point in testing this proposition was a meta- analysis designed to determine if writing and writing instruction enhanced reading (Graham & Hebert, 2011). When students in grades two to 12 wrote about material they read, there were improvements in their comprehension of the text read on researcher-designed (ES ¼ 0.50 in 55 studies) and norm-referenced measures (ES ¼ 0.37 in 11 studies). Increasing how much students wrote in grades one to six also had a positive impact on reading comprehension as measured with researcher-designed and norm-referenced tests (ES ¼ 0.35 in nine studies). When the effects of teach- ing writing (i.e., text and sentence structure instruction or process writing in 14 studies) were examined, grade four to 12 students’ reading comprehension improved on researcher-designed (ES ¼ 0.27) and norm-referenced meas- ures (ES ¼ 0.22), whereas sentence and spelling instruction improved grades one to five students’ word reading (ES ¼ 0.62 in six studies) and grades one to seven students’ read- ing fluency (ES ¼ 0.66 in five studies) on a combination of researcher-designed and norm-referenced measures. A sub- sequent meta-analysis with 20 studies examining the effects of spelling on reading (Graham & Santangelo, 2014) pro- vided further support for the impact of spelling instruction. It improved the word reading of students in kindergarten to grade six (ES ¼ 0.40 in 16 studies) as well as the reading comprehension of students in grades one to 12 (ES ¼ 0.66 in five studies).

A subsequent meta-analysis of reading and writing con- nections (Graham, Liu, Bartlett, et al., 2018) reversed the focus by examining if reading and reading instruction enhanced writing. This was the case. When preschool and secondary grade students were taught to read, they became better writers (ES ¼ 0.57 for all measures in 52 studies). Reading instruction that improved writing included teaching phonological awareness (ES ¼ 0.69 in 20 studies), phonics (ES ¼ 0.39 in 14 studies), and comprehension (ES ¼ 0.66 in 12 studies). Increasing students’ interaction with words or text through reading or observing others read also resulted in improved writing for preschool to secondary grade

students (ES ¼ 0.35 in 36 studies). Writing was improved by assigning more reading (ES ¼ 0.29 in nine studies), observing others read and react to text (ES ¼ 0.62 in eight studies), reading and analyzing the writing of others (ES ¼ 0.43 in five studies), and reading words or text to inciden- tally improve spelling (ES ¼ 0.28 in 15 studies).

A final meta-analysis on reading and writing connections (Graham, Liu, Aitken, et al., 2018) demonstrated that com- bining writing and reading instruction had a positive impact on both writing and reading. Forty-seven 47 studies were located which involved preschool to grade 12 students where no more than 60% and no less than 40% of time could be devoted to either skill in a combined program where both were taught. These programs had a positive impact on both reading (ES ¼ 0.39 for all reading measures in 33 studies) and writing (ES ¼ 0.37 for all writing measures in 37 stud- ies). Not all combined programs were effective in improving both skills though. For example, cooperative learning literacy programs and whole language did not result in statistically significant effects for writing, even though they did result in positive reading effect.

These meta-analyses provided strong support for the claim that reading is an important ingredient in learning to write, and vice versa.2 Additional research is needed, how- ever, as the amount of evidence available at each grade and different types of students is relatively limited. Further, researchers need to examine if students who experience dif- ficulty writing also evidence difficulties with reading. I expect this is the case in general because writing difficulties may reflect general learning problems, ineffective instruc- tion, or both. However, students do not have to know how to write to be able to read. As a result, reading issues may not be evident for some students experiencing difficulties learning to write.

Writing and content learning

Although learning that occurs in the content areas provides potential information for students to write about, just as reading does (Graham, Liu, Bartlett, et al., 2018), my col- leagues and I recently focused our attention on how writing can facilitate learning. Galbraith and Baaijen (2018) pro- posed two mechanisms for how writing results in learning. One, learning can be an incidental product of writing. As text is composed, writers must synthesize implicit connec- tions about content and ideas which exist in long-term memory, and they conduct additional synthesis and modifi- cation as they convert ideas to text. The extent to which writing leads to new learning depends on how the text cre- ated differs from the content held in a writer’s episodic memory. Second, learning from writing can be effortful and deliberate. When writing ideas and content retrieved from episodic memory or external sources are subjected to evalu- ation and manipulation in working memory (e.g., analysis

2The proposition that reading and writing support each other will not come as a surprise to many practitioners. For example, the Reading and Writing Project lead by Lucy Calkins at Teachers College, Columbia University, has promoted this claims for the last 35 years.

EDUCATIONAL PSYCHOLOGIST 63

and elaboration) learning can occur. These more deliberate processes transform a writer’s understanding of information by creating a more organized representation in memory.

To examine if writing about academic content facilitates learning, my colleagues and I conducted a meta-analysis which included studies where writing-to-learn was part of instruction in science, social studies, or mathematics content (Graham, Kiuhara, et al., 2020). In 56 studies conducted with students in grades one to 11, writing about academic content in these three disciplines enhanced learning (ES ¼ 0.30). Writing-to-learn was equally effective with elementary, middle, and high school students. Just as importantly, the impact of writing on learning was equivalent across the three disciplines. This finding along with the findings that writing and writing instruction improve students’ reading (Graham, Liu, Bartlett, et al., 2018), provides support for the argument that writing needs to be a more integral part of the school curriculum (Graham, 2019).

Identification of effective writing practices

Students write and learn to write in many different writing communities, including at home (Bazerman et al., 2018). One of the most important communities where writing occurs is school. The research summarized earlier about fac- tors which influence writing provides some insight into what needs to be addressed when teaching writing in school contexts. This includes increasing students’ knowledge about writing, helping them become more strategic and motivated writers, making sure they master transcription and sentence construction skills, improving their phonological awareness and depth and breadth of vocabulary, taking advantage of the reciprocal nature of reading and writing relations, using writing to support of students’ learning, and considering the role of context in students’ writing.

To obtain a more complete understanding of how to effectively teach writing to school-aged students, my col- leagues and I have applied two basic approaches to studying this issue. First, we conducted a host of intervention studies examining the effectiveness of instructional methods designed to improve specific aspects of writing such as plan- ning (e.g., Harris et al., 2006), revising (e.g., MacArthur et al., 1991), handwriting (e.g., Graham et al., 2000), spelling (e.g., Graham et al., 2002), and sentence construction (e.g., Saddler & Graham, 2007). We also tested more complete writing programs where the process approach to writing, word processing, and strategy instruction were combined (e.g., MacArthur et al., 1995). These and other investigations involved true-experiments (e.g., Graham et al., 2005) quasi- experiments (e.g., De La Paz & Graham, 2002), single partic- ipants designs (e.g., Danoff et al., 1993), mixed methods designs (e.g., McKeown et al., 2019), and case studies (e.g., MacArthur et al., 1996). Collectively, these investigations provided data for our second approach, which involved reviewing the accumulated writing intervention literature to identify effective practices.

This second approach to identifying effective writing practices included three forms of evidence. One form of

evidence involved outcomes from meta-analyses of writing interventions tested with true- and quasi-experiments (e.g., Graham & Perin, 2007). A second line of evidence included findings from a meta-analysis of single participant design research (Rogers & Graham, 2008). The final form of evi- dence focused on a meta-synthesis of qualitative studies of examining the writing practices of exceptional literacy (e.g., Graham et al., 2015).

The data from these three forms of evidence were pre- sented in two book chapters published in 2016 (Graham, Harris, et al. 2016) and 2018 (Graham & Harris, 2018). This involved a reanalysis of investigations from 20 previous reviews. My colleagues and I had conducted 14 of these reviews. In Table 1, I present outcomes from these two review of reviews. The table also contains reading and reading instructional practices which impacted writing in Graham, Liu, Aitken, et al. (2018). I added the findings from this meta- analysis to Table 1, as reading is a skill writers use when com- posing (Kim & Graham, in press) and programs that combine writing and reading instruction improve both of these skills (Graham, Liu, Bartlett, et al., 2018).

In Table 1, average weighted effects sizes are reported on the left side of the Table for instructional practices tested in three or more true- or quasi-experiments. On the right side of the Table, percent of non-overlapping data (PND) are reported for single participant design studies testing writing practices assessed in at least three investigations. In such studies, performance of each participating student is repeat- edly assessed before, during, or after instruction to establish a stable baseline against which treatment effects can be com- pared. To rule out counter explanations for any observed effects, the investigator in single participant design studies controls when instruction is introduced to each participant. PND is calculated by determining the percentage of treat- ment/after treatment data points which exceed the highest baseline score. The average percentage for all participants is the PND. A PND greater than 90% is considered a large effect; PND between 70.1% and 90% is a moderate effect; and a PND between 50.1% and 70% is a small effect. For each writing practice in Table 1, number of studies and grades of students in these investigations are reported. Unless otherwise indicated, effect sizes and PND are for quality of students’ writing. I did not include observations from the meta-synthesis of exceptional literacy teachers in Table 1, but I integrate this information into my commen- tary below.

So, what recommendations was I able to draw from the accumulated evidence? First, when students write and read, good things happen. When how much children and young adolescents write is increased, they become better writers. When young children and adolescents read, they become better writers and spellers. Writing about materials read and content presented in class enhances reading comprehension and content learning. Exceptional literacy teachers also stress the value of writing, as they dedicate time for students to write, involve students in various forms of writing, and engage students in writing across the curriculum (Graham et al., 2015).

64 S. GRAHAM

Second, teaching writing skills, processes, and knowledge is important to students’ success as writers. When students are taught foundational writing skills such as handwriting, spell- ing, and sentence construction, these skills improve as does the quality of their writing. Helping students become more thoughtful writers by teaching genre specific as well more general strategies and techniques for planning, revising, and self-regulation result in better writing. Enhancing creativity, imagery skills, and vocabulary improves what students write. Creating situations where students analyze and emulate model texts as well as observe others interact with text leads to better writing. Teaching phonological awareness and phonics strengthens spelling skills, whereas teaching reading comprehension strategies boosts the quality of students’ writing. Exceptional literacy teachers also emphasize teach- ing, as they dedicate time to writing instruction and they

teach writing strategies and foundational writing skills to the whole class, small groups, and individual students (Graham et al., 2015). When teaching writing, they model, explain, and provide guided assistance.

Third, feedback improves students’ writing. When students are provided feedback about their writing and what they are learning by their teachers, they become better writers. Teaching students how to provide constructive feedback to peers about writing results in better text. Helping students learn to assess their own writing improves text quality as does receiving feedback from a computer.

Finally, a supportive writing community is an essential element of a strong writing program. Exceptional literacy teachers place considerable emphasis on creating a writing community where students are supported and can take risks (Graham et al., 2015). They are enthusiastic about writing,

Table 1. Research-supported practices for teaching writing.

Writing methods Group studies (N) Grades Effect size Single participant design (N) Grades PND

Students write & read Extra writing 9 2–8 0.24 Extra reading 9 Pre-Sec 0.31 Reading words or text 15 Pre-Elem 0.28 (Sp) Write about reading (RDT) 55 2–12 0.50 (Rc) Write about reading (NRT) 11 2–12 0.37 (Rc) Write to Learn 31 2–12 0.29 (Lc)

Teach writing & reading Writing strategies (plan, revise, or both) 42 1�12 1.26 11 2–8 72% General strategy instruction 14 4�10 0.56 SRSD instruction 28 2–12 1.59 SIþ self-regulation 6 2–6 0.50 Strategies for editing 5 4–12 94% (Ec) Strategies for paragraph writing 4 8–9 94% (El) Creativity/imagery 4 3–6 0.76 Sentence construction 5 4–9 0.50 5 6–12 86% (Cs) Transcription skills 8 1–3 0.55 Handwriting 15

20 K–9 K–9

0.63 (Hf) 0.59 Hl)

Spelling 25 K–10 0.54 (Sp) Phonological awareness 18 Pre & Elem 0.68 (Sp) Vocabulary 3 3–8 0.78 Grammar/ usage 10 3–11 –0.17 4 2–6 83% (G) Text structure 12 2–10 0.30 Emulate model text 7 3–12 0.30 Observe readers 8 Elem-Sec 0.62 Phonics 20 Pre-Sec 0.41 (Sp) Comprehension 10 Elem-Sec 0.72

Writing assessment Adult feedback 7 2–6 0.87 Peer feedback 10 2–9 0.77 Self-assessment 13 2–12 0.51 7 2–7 51% (P0 Computer feedback 5 6–10 0.44 6þ 1 traits 4 3–6 0.05

Supportive context Process approach 33 1–12 0.34 Elementary 18 1–5 0.48 Secondary 14 6–12 0.25 Word processing 39 1–12 0.44 4 1–5 70% (p) Word processing plus 4 1–5 1.46 Collaboration 8 2–12 0.74 Writing goals 9 4–8 0.80 7 2–12 79% (P) Pre-writing 13 2–9 0.48 4 3–12 52% Inquiry 6 3–12 0.37

Reinforcement 4 3–6 96% Motivation 3 5–12 1.07

Note: Unless otherwise noted (see abbreviation in the parentheses next to an effect size or PND), the observed effect is for writing quality; all effect sizes are average-weighted effects using a randomized model; PND¼ percent of non-overlapping data; RDT¼ researcher-designed test; NRT¼ norm-referenced test; Pre¼ preschool; Elem¼ elementary grades; Sec¼ secondary grades; Sp¼ spelling; Rc¼ reading comprehension; Lc¼ learning content; Ec¼ errors corrected; El¼ genre elements; Cs¼ correct sentences; K¼ kindergarten; Hf¼ handwriting fluency; Hl¼ handwriting legibility; G¼ grammar; P¼ production (writing out- put); effect sizes for teaching grammar and text-structure as well as the 6þ 1 approach were not statistically greater than no effect; PND greater than 90% is a large effect; PND between 70.1% and 90% is a moderate effect; PND 50.1% to 70% is a small effect.

EDUCATIONAL PSYCHOLOGIST 65

set high expectations for their students, encourage students to do their best, emphasize the skills and processes taught are essential to good writing, attribute writing success to effort, engage students in thoughtful activities, encourage students to act in a self-regulated manner, adapt writing assignments and instruction to meet students’ needs, and establish routines where students plan, draft, edit, revise, and share their writing with others.

The outcomes from the quantitative studies summarized in Table 1 provide empirical support for many of the actions of the exceptional literacy teachers described above and extends them. For example, establishing routines for writing as the process approach method does results in better writ- ing as do procedures designed to enhance motivation for writing, including reinforcing students’ writing efforts. Establishing specific goals for writing increases the product- ivity and quality of students’ writing. Having students com- pose together improves what each students writes when composing alone. Helping students gather, think about, and organize ideas for writing through inquiry or prewriting activities leads to better writing. Making word processing the primary tool for composing in the classroom enhances the quality of students’ writing. This is especially the case for word processing programs which provide students with supports which extend beyond spelling and gram- mar checkers.

My colleagues and I have also conducted meta-analyses to determine what practices are effective with students with disabilities (Gillespie & Graham, 2014; Rogers & Graham, 2008). Many of the same practices that are effective with students in general are effective with these students. This includes teaching writing strategies, sentence construction skills, and handwriting and spelling. It also includes the pro- cess approach to writing, establishing goals for writing, word processing, pre-writing activities, and self-monitoring. The primary difference in studies conducted with students with disabilities versus studies in the general school population involves providing students with disability individual or small group instruction, mastery learning, and more intense instruction.

Knowledge of writing and how to teach it has become increasingly informed by research. There is still much to learn though. Just as importantly, there is much that still needs to be done to bring this knowledge into the classroom and for teachers to use it effectively.

Writing instruction in schools

Writing is an extremely versatile skill. As I already estab- lished, writing and writing instruction make students better readers (Graham & Hebert, 2011), and writing about mater- ial read or presented in class facilitates learning (Graham, Kiuhara, et al., 2020). People also use writing to communi- cate, persuade, inform, record information, chronicle experi- ences, think about topics, express feelings, create imaginary worlds, entertain others, explore the meaning of events, and heal psychological and physiological wounds (Bazerman et al., 2018). Given the power of writing and its complexity,

it seems likely that schools would devote considerable effort and time to teaching it. My colleagues and I have examined this proposition using survey methods, including asking teachers in the United States (U.S.; Brindle et al., 2016; Cutler & Graham, 2008; Gilbert & Graham, 2010; Kiuhara et al., 2009; Ray et al., 2016), the Greater China region (Hsiang et al., 2018, 2020), Europe (Graham et al., 2021), and South America (Ba~nales et al., 2020) how they teach writing. In all of the studies conducted in the U.S. and the Greater China region, teachers were randomly selected from the population of teachers.

The findings from these and other surveys and observa- tions of teachers’ writing practices from around the world are summarized in Graham (2019). This analysis of 28 stud- ies conducted since 2000 (14 studies were conducted by my colleagues and I) yielded two basic findings. On the one hand, some teachers provided solid and even exemplary writing instruction, devoting considerable time to teaching writing and frequently applying research-supported practi- ces. On the other hand, most teachers in these studies did not provide their students with a solid writing program, as students did little writing in their classrooms and little time was spent teaching writing. Although these teachers com- monly applied a variety of research-based writing practices, they applied them infrequently. They also made frequent adaptations for their weakest writers, but a sizeable minority of teachers in most studies made few or no adaptations. Unfortunately, these findings were generally universal, applying across different grades and countries.

Given the importance of writing and the common view that teaching writing is an important responsibility of schools, it is hard to imagine how schools, teachers, and stu- dents ended up in this particular cull de sac. There are many reasons for why this may be the case though, includ- ing societal beliefs that reading instruction is all that is needed to become a good writer (Hsiang & Graham, 2016), good writers are born and not made (Hsiang et al., 2020), writing is not as important as other school subjects (Hsiang et al., 2018), and writing is acquired naturally much like oral language (Graham & Harris, 1977). The teaching of writing can also be constrained by institutional and political decisions. For example, in the Netherlands, teachers can meet the expectations established by the Dutch Inspectorate for writing by teaching just two writing lessons a month (Rietdijk et al., 2018). As a further example, the teaching of writing receives very little emphasis in teacher preparation programs in the U.S. (Graham, 2019).

Teachers’ beliefs also play an important role in the teach- ing of writing. To illustrate, teachers who believe they are better prepared to teach writing, are more confident writing teachers, derive greater pleasure from teaching writing, and place greater value on its importance, are more likely to devote more time and effort to teaching this complex skill (Brindle et al., 2016; Gilbert & Graham, 2010; Hsiang & Graham, 2016; Kiuhara et al., 2009). As a result, any effort to change how writing is taught must take into account teachers’ beliefs and motivations regarding writing.

66 S. GRAHAM

If writing is to become a more central aspect of school- ing, it must become more valued and its importance better understood by society at large, policymakers, and school personnel (Graham, 2019). Better writing instruction depends on systemic changes which involve public percep- tions, rock solid commitments by political and educational systems to its value, and teachers who are prepared and want to teach it. The knowledge to make this happen exists, but it is not clear if there is the will to do so. Given the ver- satility and power of writing, I hope the current story line which dominates the teaching of writing in most schools can be flipped, and writing becomes the valued tool it should be.

Concluding comments

Although my colleagues and I have made what we hope are important contributions to the field of writing and educa- tional psychology, there is still much work to be done. Our knowledge of how writing operates, evolves, and is taught effectively is incomplete. Too little research has focused on the connections between the communities where writing occurs and is learned, the cognitive resources and capabil- ities that members of these communities (writers, collabora- tors, teachers, mentors, and readers) bring to bear when writing or learning to write, how writing is acquired and taught in various writing communities, and how broader societal, cultural, institutional, political, and historical forces shape and bound writing communities and their constitu- ents (Graham, 2018a). Scholars are just beginning to under- stand how writing develops similarly and differently across distinct languages and writing systems. Our knowledge about writing in the workplace and at home as well as how people grow as writers across the lifespan is underdeveloped. There are plenty of new research opportunities available so that young scholars can engage in a rich, productive, and impactful program of research in writing.

Writing growth

I want to note that I used the term growth and not develop- ment to refer to improvements that occurred in students’ writing. I do not believe that writing develops naturally (Graham & Harris, 1997). Although there may be various stages that writers go through as they grow and become more skilled writers over the lifespan, our understanding of these stages is quite limited at this point, except in areas such as early spelling (Treiman, 1997) and children’s earliest attempts at writing (see Graham & Weintraub, 1996 for examples). Instead, I have come to think of students’ growth as writers as an increasing potential to write effectively for a broadening multitude of purposes and in an increasing array of contexts (see Bazerman et al., 2018). Growth occurs as students accumulate and integrate new understandings about writing as well as new writing beliefs, skills, processes, and strategies, resulting in a reorganization or re-alignment of their capabilities to apply and control the medium of writing and its meaning-making practices. Such changes may be

quite restricted and apply to a specific type of writing in a particular situation (e.g., developing a new understanding about the importance of evaluating multiple sides of an argument when writing a persuasive paper in science) or more expansive applying to writing more broadly (e.g., learning the value of reading aloud written papers to iden- tify grammar or other writing miscues).

My view of growth in writing is also colored by the spec- ter of variability. There is considerable variability between and within students in their capabilities and desires to write for a variety of purposes and in different contexts. Just because a child can consistently write an engaging story does not mean she or he can consistently write a clear and persuasive text that justifies a solution to a particular prob- lem (Graham et al., 2016). For example, I am pretty good at writing an article for a journal, but pretty much a novice at writing a novel, novella, or short story for others to read. Writing is not a single thing. This is made even more com- plicated, as the purposes, audiences, norms, values, social practices, writing tools, motivations, typical practices can differ from one writing community to the next (e.g., home, school, on-line).

Differences in context and what they mean for students’ growth as writers can be illustrated by considering how the purposes and even the fabric of writing tasks differ between China and the U.S. (Hsiang et al., 2018). Teachers in China traditionally view writing as a tool for shaping and educat- ing students’ minds, whereas teachers in the U.S. tend to view writing as a tool for self-disclosure and self-expression (Li, 1996). Moreover, writing an argument involves a less direct approach in China than it does in the U.S. class- rooms. Instead of directly stating their thesis, Chinese stu- dents are encouraged to imply the intended message and to rely on clich�es and set phrases to foster group values and avoid conflict. What is considered acceptable growth in writ- ing in these two contexts is likely to differ.

This does not say nothing is known about how students grow and prosper as writers, as I hope this article made clear. For example, the mastery of some skills, like handwrit- ing and spelling, in the primary grades yields positive bene- fits to students’ growth as writers. Teaching students how to plan, evaluate, monitor and revise what they write can begin as early as first grade and should continue until graduation, with students learning increasingly sophisticated strategies for carrying out these processes. Students need to become progressively skilled at translating their thought into cor- rectly written sentences that captures their intentions as well as the attention of their readers. Students need to acquire a diverse, integrated, and expanding arrays of knowledge about the purposes and characteristics of different types of writing, so that they can create such text but also go beyond them. Students need to write. These five points are likely valid across different cultures and orthographies (e.g., alpha- betic, logographic) in my opinion.

These five claims do not capture all that is known about how to help students grow as writers, but it is important to note they are general in their specifications. This reflects the state of our current understanding of writing based on

EDUCATIONAL PSYCHOLOGIST 67

accumulated research evidence gathered at this point and time. In essence, it is not possible to put forth research- based benchmarks or norms for what students can and should be able to master at each grade. Attempts to do so, such as the Common Core State Standards (2010), are edu- cated guesses (Graham et al., 2015).

Assuming it is possible in the future to establish research-based norms for writing growth, care must be exer- cised when applying them. As abstract milestones, research- as well as socially-derived benchmarks do not capture the heterogeneity in students’ writing, the variability in the processes and timing by which children reach them, or dif- ferences in students’ performance depending on the pur- poses and contexts of writing (Bazerman et al., 2018). Such milestones may inadvertently lead educators to believe that there is a single path and endpoint to writing growth. Instead, writing trajectories are neither singular nor are they always forward moving. For example, backward transitions are possible when students use less sophisticated strategies than normal when confronted with unfamiliar writing tasks or writing in an unfamiliar social situation (Bazerman et al., 2017).

It must further be recognized that research- or socially- derived milestones can underestimate as well as overestimate what students can do in a given situation. For instance, a relatively consistent finding in the research literature was that revising had little effect on students’ writing until high school (Fitzgerald, 1987). However, it would be a mistake to develop writing benchmarks based on this finding, as researchers also demonstrated that younger students, even those who find writing challenging, can improve the quality of their writing through revising when they are taught how to revise effectively (De La Paz et al., 1998). What students can and cannot do when writing is not a fixed target, as it is directly influenced by the instruction and scaffolding stu- dents receive. I am not saying that benchmarks are not use- ful. Rather, a healthy dose of skepticism must be applied when putting them into action.

Consequently, the field has come a long way in what is known about writing growth and the teaching of writing. There is still a long way to go though. At this point, I believe the best course of action is to take an evidence-based practice (EBP) approach to teaching writing. The EBP movement is based on the idea that practitioners should apply the best sci- entific evidence available to make informed and judicious decisions for their clients (Sackett et al., 1996). This does not mean that teachers should abandon the knowledge about teaching writing they have acquired through their experiences as teachers or learners. Rather, the EBP movement stresses that teachers apply what has been learned from science by contextualizing this knowledge with their own knowledge about their students, the context in which they work, and what they know about writing and teaching it.

New writing technologies

It is also important to realize how people write is rapidly changing. New digital systems and writing tools provide an

array of options for creating text. The development of the internet and digital-based writing tools makes it easier to: (a) create and share text with others; (b) access content for our writing through digitally stored information; (c) produce text that includes sound, talk, drawings, animations, film, and links to other material; (d) generate written text by speaking aloud; and (e) receive feedback and other forms of support from the very machines used for writing (Graham, in press-b). In the very near future, it is conceivable that machines will serve as our writing partners, composing text with us.

The advent of these new tools for writing has far reach- ing implications for writing and the teaching of writing. In a relatively short period of time, they have provided new ways of generating text (e.g., speech-to-text synthesis) new mediums for sharing text (e.g., electronically), new modes for presenting the message (e.g., multi-modal text), new means for gathering ideas for writing (e.g., accessing infor- mation stored on the web), new opportunities for writing collaboratively (e.g., blogs, wikis, and RSS feeds to facilitate writing collaboration), and new means of support (e.g., spell and grammar checkers). There are now automated essay evaluation technologies, apps, and computer-based instruc- tional procedures for supporting and teaching writing (see Allen et al., 2016; Boucher et al., 2017; Wijekumar et al., 2017).

These new writing technologies and the ones that will appear in the near future are challenging conceptualizations of what it means to write (Graham, in press-b). The trad- itional view of writing at school involves each student writ- ing alone, creating a text for a very limited audience (the writer, teacher, and possibly classmates). New writing tech- nologies make it easier for students to work together, as multiple youngsters can gather around a computer screen to compose collaboratively, students can access and edit each other’s writing through an integrated digital network, and students can access an online Wiki to work collectively on a writing project. Likewise, what students can write can be shared broadly via email or by posting it on the web.

Further, the development of multimodal writing systems allows students to create products which include text, audio, images, animation, and film. This challenges the traditional view of writing as printed text. These changes, including the role of machines as collaborators in the writing process in the foreseeable future, is the leading edge of a revolution that is coming more clearly into view (Graham, in press-b). Even so, it is not clear how writing will be defined as these tools become more common in the classroom and what will count as writing. For example, if a poem includes only sounds and images, but no words, is it writing?

An especially difficult dilemma for schools is that they cannot abandon older writing technologies such as paper, pencil, and pen. These writing tools are widely used at home, school, and work (Lenhart et al., 2008), and they are not likely to disappear any time soon because they are cheap, durable, and easily transportable (Graham, in press- b). However, if schools are to remain relevant, they must integrate some of these new writing technologies into their

68 S. GRAHAM

curriculum, so that students can learn to master these devi- ces and use them effectively in academic settings. Making this challenge even more complex is that new writing tools are created at a dizzying rate (Leu et al., 2016). This raises multiple questions, including: What writing tools should stu- dent learn to use at school? How does the introduction of new writing technologies at school change how writing is defined and taught? How can these new writing technologies be used to support reading and learning? Can the introduc- tion of new writing technologies in schools be done so that it is equitable and fair?

For those of us who study writing and teach it, this changing landscape of writing presents new challenges and opportunities. How schools adjust to the availability of these new tools and how students use them will depend on mul- tiple intervening variables such as economic development, ideology, history, ethnic and linguistic composition, social structure, and institutional factors (Graham, in press-b). I hope the research summarized here provides new insight into the value of writing and the dazzling array of possibil- ities for studying it. It is an exhilarating time to be a writing researcher, and I am excited to see what comes next.

Funding

The completion of this article was supported from the Institute of Education Sciences by a grant [R305C190007] to the WRITE (Writing Research to Improve Teaching and Evaluation) Center for Secondary Students, University of California, Irvine.

References

Alexander, P. (1998). The nature of disciplinary and domain learning: The knowledge, interest, and strategic dimensions of learning from subject-matter text. In C. Hynd (Ed.), Learning from text across con- ceptual domains (pp. 55–76). Erlbaum.

Allen, L., Jacovina, M., & McNamara, D. (2016). Computer-based writ- ing instruction. In C. A. MacArthur, S. Graham, & J. Fitzgerald (Eds.), Handbook of writing research (2nd ed., pp. 330–347). Guilford Press.

Ba~nales, G., Ahumada, S., Graham, S., Puente, A., Guajardo, M., & Mu~noz, I. (2020). Teaching writing in grades 4–6 in Urban Schools in chile: A National Survey. Reading and Writing, 33(10), 2661–2696. https://doi.org/10.1007/s11145-020-10055-z

Bangert-Drowns, R. L., Hurley, M. M., & Wilkinson, B. (2004). The effects of school-based Writing-to-learn interventions on academic achievement: A meta-analysis. Review of Educational Research, 74(1), 29–58. https://doi.org/10.3102/00346543074001029

Bazerman, C., Applebee, A., Berninger, V., Brandt, D., Graham, S., Matsuda, P., Murphy, S., Rowe, D., & Schleppegrell, M. (2017). Taking the long view on writing development. Research in the Teaching of English, 51(3), 351–360. https://library.ncte.org/jour- nals/rte/issues/v51-3/28980

Bazerman, C., Berninger, V., Brandt, D., Graham, S., Langer, J., Murphy, S., Matsuda, P., Rowe, D., & Schleppegrell, M. (2018). The lifespan development of writing. National Council of English.

Bereiter, C., & Scardamalia, M. (1982). From conversation to compos- ition. In R. Glaser (Ed.), Advances in instructional psychology. (Vol. 2, pp. 1–64). Erlbaum.

Berninger, V. W. (1999). Coordinating transcription and text gener- ation in working memory during composing: Automatic and con- structive processes. Learning Disability Quarterly, 22(2), 99–112. https://doi.org/10.2307/1511269

Berninger, V., Vaughan, K., Abbott, R., Abbott, S., Rogan, L., Brooks, A., Reed, E., & Graham, S. (1997). Treatment of handwriting prob- lems in beginning writers: Transfer from handwriting to compos- ition. Journal of Educational Psychology, 89(4), 652–666. https://doi. org/10.1037/0022-0663.89.4.652

Berninger, V. W., Vaughan, K., Abbott, R. D., Brooks, A., Abbott, S. P., Rogan, L., Reed, E., & Graham, S. (1998). A multiple connec- tions approach to early intervention for spelling problems: Integrating instructional, learner and stimulus variables. Journal of Educational Psychology, 90(4), 587–605. https://doi.org/10.1037/ 0022-0663.90.4.587

Boucher, A., Evans, M., & Graham, S. (2017). Udio: Rich and authentic literacy experiences for struggling middle school readers. In S. Crossley & D. McNamara (Eds.), Adaptive Educational Technologies for Literacy. (pp. 294–301). Taylor & Francis.

Brindle, M., Harris, K. R., Graham, S., & Hebert, M. (2016). Third and fourth grade teachers’ classroom practices in writing: A national survey. Reading and Writing, 29(5), 929–954. https://doi.org/10. 1007/s11145-015-9604-x

Camping, A., Graham, S., Ng, C., Aitken, A., Wilson, J., & Wdowin, J. (2020). Writing Motivation of middle school emergent bilingual stu- dents. Reading and Writing, 33(9), 2361–2390. https://doi.org/10. 1007/s11145-020-10046-0

Common Core State Standards: National Governors Association and Council of Chief School Officers (2010). Retrieved from http://www. corestandards.org/.

Cutler, L., & Graham, S. (2008). Primary grade writing instruction: A national survey. Journal of Educational Psychology, 100(4), 907–919. https://doi.org/10.1037/a0012656

Danoff, B., Harris, K. R., & Graham, S. (1993). Incorporating strategy instruction into the school curriculum: Effects on children’s writing. Journal of Reading Behavior, 25(3), 295–322. https://doi.org/10.1080/ 10862969009547819

De La Paz, S., & Graham, S. (2002). Explicitly teaching strategies, skills, and knowledge: Writing instruction in middle school classrooms. Journal of Educational Psychology, 94(4), 687–698. https://doi.org/10. 1037/0022-0663.94.4.687

De La Paz, S., Swanson, P., & Graham, S. (1998). Contribution of executive control to the revising problems of students with writing and learning difficulties. Journal of Educational Psychology, 90(3), 448–460. https://doi.org/10.1037/0022-0663.90.3.448

De Smedt, F., Van Keer, H., & Graham, S. (2019). The “bright” and “dark” side of writing motivation: Effects of explicit instruction and writing with peers. The Journal of Educational Research, 112(2), 152–167. https://doi.org/10.1080/00220671.2018.1461598

Fitzgerald, J. (1987). Research on revision in writing. Review of Educational Research, 57(4), 481–506. https://doi.org/10.3102/ 00346543057004481

Galbraith, D., & Baaijen, V. (2018). The work of writing: Raiding the inarticulate. Educational Psychologist, 53(4), 238–257. https://doi.org/ 10.1080/00461520.2018.1505515

Gilbert, J., & Graham, S. (2010). Teaching writing to elementary stu- dents in grades 4 to 6: A national survey. The Elementary School Journal, 110(4), 494–518. https://doi.org/10.1086/651193

Gillespie, A., & Graham, S. (2014). A meta-analysis of writing interven- tions for students with learning disabilities. Exceptional Children, 80(4), 454–473. https://doi.org/10.1177/0014402914527238

Gillespie, A., Graham, S., & Compton, D. (2017). Writing to learn in science. The Journal of Educational Research, 110(4), 366–379. https://doi.org/10.1080/00220671.2015.1103688

Graham, S. (1982). Written composition research and practice: A uni- fied approach. Focus on Exceptional Children, 14(8), 1–16. https:// doi.org/10.17161/foec.v14i8.7447

Graham, S. (1983). The effects of a self-instructional procedure on learning disabilities students’ handwriting performance. Learning Disability Quarterly, 6(2), 231–234. https://doi.org/10.2307/1510802

Graham, S. (1990). The role of production factors in learning disabled students’ compositions. Journal of Educational Psychology, 82(4), 781–791. https://doi.org/10.1037/0022-0663.82.4.781

EDUCATIONAL PSYCHOLOGIST 69

Graham, S. (2006). Writing. In P. Alexander & P. Winne (Eds.), Handbook of educational psychology (pp. 457–478). Erlbaum.

Graham, S. (2018a). The writer(s)-within-community model of writing. Educational Psychologist, 53(4), 258–279. https://doi.org/10.1080/ 00461520.2018.1481406

Graham, S. (2018b). A writer(s) within community model of writing. In C. Bazerman, V. Berninger, D. Brandt, S. Graham, J. Langer, S. Murphy, P. Matsuda, D. Rowe, & M. Schleppegrell,(Eds.), The life- span development of writing (pp. 271–325). NCTE.

Graham, S. (2019). Changing how writing is taught. Review of Research in Education, 43(1), 277–303. https://doi.org/10.3102/ 0091732X18821125

Graham, S. (2020). The sciences of reading and writing must become more fully integrated. Reading Research Quarterly, 55 (S1), S35–S44. https://doi.org/10.1002/rrq.332

Graham, S. (2021). Through the looking glass: Reflections of a writing researcher. Education Review, 28, 28. https://edrev.asu.edu/index. php/ER/article/view/3209

Graham, S. (in press-a). Writer(s)-Within-Community model of writ- ing as a lens for studying the teaching of writing. In R. Horrowitz (Ed.), The Routledge handbook of international research on writing, Vol. II. Routledge.

Graham, S. (in press-b). Teaching writing in the digital age. In T. L. Good, & M. McCaslin (Eds.), Routledge encyclopedia of education (Online). Taylor & Francis.

Graham, S., Aitken, A., Hebert, M., Santangelo, T., Camping, A., Harris, K. R., Eustice, K., & Ng, C. (2020). Do children with reading difficulties experience writing difficulties? A meta-analysis. Journal of Educational Psychology, 2020, 643. https://doi.org/10.1037/ edu0000643

Graham, S., Berninger, V., Abbott, R., Abbott, S., & Whitaker, D. (1997). The role of mechanics in composing of elementary school students: A new methodological approach. Journal of Educational Psychology, 89(1), 170–182. https://doi.org/10.1037/0022-0663.89.1. 170

Graham, S., Berninger, V., Weintraub, N., & Schafer, W. (1998). The development of handwriting fluency and legibility in grades l through 9. The Journal of Educational Research, 92(1), 42–52. https://doi.org/10.1080/00220679809597574

Graham, S., Collins, A., & Rigby-Wills, H. (2017). A meta-analysis examining the writing characteristics of students with learning dis- abilities and normally achieving peers. Exceptional Children, 83(2), 199–218. https://doi.org/10.1177/0014402916664070

Graham, S., Daley, S., Aitken, A., Harris, K. R., & Robinson, K. (2018). Do writing motivational beliefs predict middle school students’ writ- ing performance? Journal of Research in Reading, 41(4), 642–656. https://doi.org/10.1111/1467-9817.12245

Graham, S., Fishman, E., Reid, R., & Hebert, M. (2016). Writing char- acteristics of students with ADHD and their normally achieving peers. Learning Disabilities Research & Practice, 31(2), 75–89. https://doi.org/10.1111/ldrp.12099

Graham, S., Harbaugh-Schattenkirk, A., Aitken, A., Harris, K. R., Nj, C., Wilson, J., & Wdowin, J. (2021). Writing motivation question- naire: Validation and application as a formative assessment. Submitted for publication.

Graham, S., & Harris, K. R. (1989). A components analysis of cognitive strategy instruction: Effects on learning disabled students’ composi- tions and self-efficacy. Journal of Educational Psychology, 81(3), 353–361. https://doi.org/10.1037/0022-0663.81.3.353

Graham, S., & Harris, K. R. (1997). It can be taught, but it does not develop naturally: Myths and realities in writing instruction. School Psychology Review, 26(3), 414–424. https://doi.org/10.1080/02796015. 1997.12085875

Graham, S., & Harris, K. R. (2000). The role of self-regulation and transcription skills in writing and writing development. Educational Psychologist, 35(1), 3–12. https://doi.org/10.1207/ S15326985EP3501_2

Graham, S., & Harris, K. R. (2018). Evidence-based writing practices: A meta-analysis of existing meta-analyses. In R. Fidalgo, K. R. Harris, & M. Braaksma (Eds.), Design principles for teaching effective writing:

Theoretical and empirical grounded principles (pp. 13–37). Brill Editions.

Graham, S., Harris, K. R., & Adkins, M. (2018). The impact of supple- mental handwriting and spelling instruction with first grade students who do not acquire transcription skills as rapidly as peers: A randomized control trial. Reading and Writing, 34(6), 1273–1294. https://doi.org/10.1007/s11145-018-9822-0

Graham, S., Harris, K. R., & Chambers, A. (2016). Evidence-based practice and writing instruction. In C. MacArthur, S. Graham, & J. Fitzgerald (Eds.), Handbook of writing research (Vol. 2; p. 211–226). Guilford.

Graham, S., Harris, K. R., & Fink, B. (2000). Is handwriting causally related to learning to write? Treatment of handwriting problems in beginning writers. Journal of Educational Psychology, 92(4), 620–633. https://doi.org/10.1037/0022-0663.92.4.620

Graham, S., Harris, K. R., & Fink-Chorzempa, B. (2002). Contributions of spelling instruction to the spelling, writing, and reading of poor spellers. Journal of Educational Psychology, 94(4), 669–686. https:// doi.org/10.1037/0022-0663.94.4.669

Graham, S., Harris, K. R., & Mason, L. (2005). Improving the writing performance, knowledge, and motivation of struggling young writ- ers: The effects of self-regulated strategy development. Contemporary Educational Psychology, 30(2), 207–241. https://doi.org/10.1016/j.ced- psych.2004.08.001

Graham, S., Harris, K. R., & McKeown, D. (2013). The writing of stu- dents with LD and a meta-analysis of SRSD writing intervention studies: Redux. In L. Swanson, K. R. Harris, & S. Graham (Eds.), Handbook of learning disabilities (2nd ed., pp. 405–438). Guilford Press.

Graham, S., Harris, K. R., & Santangelo, T. (2015). Research-based writing practices and the common core: Meta-analysis and meta- synthesis. The Elementary School Journal, 115(4), 498–522. https:// doi.org/10.1086/681964

Graham, S., & Hebert, M. (2011). Writing-to-read: A meta-analysis of the impact of writing and writing instruction on reading. Harvard Educational Review, 81(4), 710–744. https://doi.org/10.17763/haer.81. 4.t2k0m13756113566

Graham, S., Hebert, M., Fishman, E., Ray, A., & Gillespie-Rouse, A. (2020). Do children classified with specific language impairment Have a learning disability in writing? A meta-analysis. Journal of Learning Disabilities, 53(4), 292–310. https://doi.org/10.1177/ 0022219420917338

Graham, S., Hebert, M., Sandbank, M., & Harris, K. R. (2016). Credibly assessing the writing achievement of young struggling writ- ers: Application of generalizability theory. Learning Disability Quarterly, 39(2), 72–82. https://doi.org/10.1177/0731948714555019

Graham, S., Kiuhara, S., Harris, K. R., & Fishman, E. (2017). The rela- tionship between strategic behavior, motivation, and writing per- formance with young, developing writers. The Elementary School Journal, 118(1), 82–104. https://doi.org/10.1086/693009

Graham, S., Kiuhara, S., & MacKay, M. (2020). The effects of writing on learning in science, social studies, and mathematics: A meta-ana- lysis. Review of Educational Research, 90(2), 179–226. https://doi.org/ 10.3102/0034654320914744

Graham, S., Kiuhara, S., McKeown, D., & Harris, K. R. (2012). A meta- analysis of writing instruction for students in the elementary grades. Journal of Educational Psychology, 104(4), 879–896. https://doi.org/ 10.1037/a0029185

Graham, S., Liu, K., Bartlett, B., Ng, C., Harris, K. R., Aitken, A., Barkel, A., Kavanaugh, C., & Talukdar, J. (2018). Reading for writ- ing: A meta-analysis of the impact of reading and reading instruc- tion on writing. Review of Educational Research, 88(2), 243–284. https://doi.org/10.3102/0034654317746927

Graham, S., Liu, X., Aitken, A., Ng, C., Bartlett, B., Harris, K. R., & Holzapfel, J. (2018). Balancing reading and writing instruction: A meta-analysis. Reading Research Quarterly, 53(3), 279–304. https:// doi.org/10.1002/rrq.194

Graham, S., & Miller, L. (1979). Spelling research and practice: A uni- fied approach. Focus on Exceptional Children, 12(2), 1–16. https:// doi.org/10.17161/foec.v12i2.7150

70 S. GRAHAM

Graham, S., & Perin, D. (2007). A meta-analysis of writing instruction for adolescent students. Journal of Educational Psychology, 99(3), 445–476. https://doi.org/10.1037/0022-0663.99.3.445

Graham, S., & Santangelo, T. (2014). Does spelling instruction make students better spellers, readers, and writers? A meta-analytic review. Reading and Writing, 27(9), 1703–1743. https://doi.org/10.1007/ s11145-014-9517-0

Graham, S., Schwartz, S., & MacArthur, C. (1993). Knowledge of writ- ing and the composing process, attitude toward writing, and self- efficacy for students with and without learning disabilities. Journal of Learning Disabilities, 26(4), 237–249. https://doi.org/10.1177/ 002221949302600404

Graham, S., Skar, G., & Falk, D. (2021). Teaching writing in the pri- mary grades in Norway: A national survey. Reading and Writing, 34(2), 529–569. https://doi.org/10.1007/s11145-020-10080-y

Graham, S., & Weintraub, N. (1996). A review of handwriting research: Progress and prospect from l980 to l993. Educational Psychology Review, 8(1), 7–87. https://doi.org/10.1007/BF01761831

Graham, S., Wijekumar, K., Harris, K. R., Lei, P., Fishman, E., Ray, A., & Houston, J. (2019). Writing skills, knowledge, motivation, and strategic behavior predict students’ persuasive writing performance in the context of robust writing instruction. The Elementary School Journal, 119(3), 487–510. https://doi.org/10.1086/701720

Harris, K. R. (1982). Cognitive behavior modification: Application with exceptional students. Focus on Exceptional Children, 15(2), 1–16. https://doi.org/10.17161/foec.v15i2.7449

Harris, K. R., Danoff-Friedlander, B., Saddler, B., Frizzelle, R., & Graham, S. (2005). Self-monitoring of attention versus self-monitor- ing of academic performance: Differential effects among students with ADHD in the regular classroom. The Journal of Special Education, 39(3), 145–156. https://doi.org/10.1177/00224669050390030201

Harris, K., & Graham, S. (1985). Improving learning disabled students’ composition skills: Self-control strategy training. Learning Disability Quarterly, 8(1), 27–36. https://doi.org/10.2307/1510905

Harris, K. R., & Graham, S. (2009). Self-regulated strategy development in writing: Premises, evolution, and the future. British Journal of Educational Psychology, 2(6), 113–135. https://doi.org/10.1348/ 978185409X422542

Harris, K. R., Graham, S., & Mason, L. (2006). Improving the writing, knowledge, and motivation of struggling young writers: Effects of Self-Regulated Strategy development with and without peer support. American Educational Research Journal, 43(2), 295–340. https://doi. org/10.3102/00028312043002295

Harris, K. R., Lane, K., Graham, S., Driscoll, S., Sandmel, K., Brindle, M., & Schatschneider, C. (2012). Practice-based professional devel- opment for strategies instruction in writing: A randomized con- trolled study. Journal of Teacher Education, 63(2), 103–119. https:// doi.org/10.1177/0022487111429005

Harris, K. R., Ray, A., Graham, S., & Houston, J. (2019). Answering the challenge: SRSD instruction for close reading of text for writing persuasively with 4th and 5th grade students experiencing writing difficulties. Reading and Writing, 32(6), 1459–1482. https://doi.org/ 10.1007/s11145-018-9910-1

Hayes, J. (1996). A new framework for understanding cognition and affect in writing. In M. Levy & S. Ransdell (Eds.), The science of writing: Theories, methods, individual differences, and applications (pp. 1–27). Erbaum.

Hebert, M., Graham, S., Rigby-Wills, H., & Ganson, K. (2014). Effects of note-taking and extended writing on expository text: Who bene- fits? Learning Disabilities: A Contemporary Journal, 12(1), 43–68.

Hsiang, T., & Graham, S. (2016). Teaching writing in grades 4 to 6 in urban schools in the Greater China region. Reading and Writing, 29(5), 869–902. https://doi.org/10.1007/s11145-015-9597-5

Hsiang, T., Graham, S., & Wong, P. (2018). Teaching writing in grades 7–9 in urban schools in the Greater China Region. Reading Research Quarterly, 53(4), 473–507. https://doi.org/10.1002/rrq.213

Hsiang, T., Graham, S., & Yang, Y. (2020). Teachers’ practices and beliefs about teaching writing: A national survey of grades 1 to 3 teachers. Reading and Writing, 33(10), 2511–2548. https://doi.org/10. 1007/s11145-020-10050-4

Kim, Y., & Graham, S. (in press). Expanding the direct and indirect effects model of writing (DIEW): Dynamic relations of component skills to writing quality, productivity, fluency. Journal of Educational Psychology.

Kiuhara, S., Graham, S., & Hawken, L. (2009). Teaching writing to high school students: A national survey. Journal of Educational Psychology, 101(1), 136–160. https://doi.org/10.1037/a0013097

Lenhart, A., Arafeh, S., Smith, A., & Macgill, A. (2008). Writing, tech- nology and teens. Pew Internet & American Life Project.

Leu, D., Slomp, D., Zawilinski, L., & Corrigan, J. (2016). Writing research through a new literacies lens. In C. A. MacArthur, S. Graham, & J. Fitzgerald (Eds.), Handbook of writing research (2nd ed., pp. 41–56). Guilford Press.

Li, X. (1996). “Good writing” in cross cultural contexts. State University of New York Press.

Limpo, T., Filipe, M., Magalhaes, S., Cordeiro, C., Veloso, A., Castro, S., & Graham, S. (2020). Development and validation of instruments to measure Portuguese third graders' reasons to write and self-effi- cacy. Reading and Writing, 33(9), 2173–2204. https://doi.org/10. 1007/s11145-020-10039-z

MacArthur, C. A., Graham, S., Schwartz, S. S., & Schafer, W. D. (1995). Evaluation of a writing instruction model that integrated a process approach, strategy instruction, and word processing. Learning Disability Quarterly, 18(4), 278–291. https://doi.org/10. 2307/1511234

MacArthur, C., Schwartz, S., & Graham, S. (1991). Effect of a recipro- cal peer revision strategy in special education classrooms. Learning Disabilities Research and Practice, 6(4), 201–210.

MacArthur, C., Schwartz, S., Graham, S., Molloy, D., & Harris, K. R. (1996). Integration of strategy instruction into a whole language classroom: A case study. Learning Disabilities Research and Practice, 11(3), 168–176.

McKeown, D., Brindle, M., Harris, K. R., Sandmel, K., Steinbrecher, T., Graham, S., Lane, K., & Oakes, W. (2019). Teachers’ voices: Perceptions of effective professional development and classwide implementation of self-regulated strategy development in writing. American Educational Research Journal, 56(3), 753–791. https://doi. org/10.3102/0002831218804146

Meichenbaum, D. (1977). Cognitive behavior modification: An integra- tive approach. Plenum.

Olinghouse, N., & Graham, S. (2009). The relationship between the writing knowledge and the writing performance of elementary-grade students. Journal of Educational Psychology, 101(1), 37–50. https:// doi.org/10.1037/a0013462

Olinghouse, N., Graham, S., & Gillespie, A. (2015). The relationship of discourse and topic knowledge to writing performance. Journal of Educational Psychology, 107(2), 391–406. https://doi.org/10.1037/ a0037549

Page-Voth, V., & Graham, S. (1999). Effects of goal setting and strategy use on the writing performance and self-efficacy of students with writing and learning problems. Journal of Educational Psychology, 91(2), 230–240. https://doi.org/10.1037/0022-0663.91.2.230

Plimpton, G. (1989). The writer’s chapbook. Modern Library. Proctor, P., Daley, S., Xu, L., Graham, S., & Hall. T. (2020). Shared

knowledge between reading and writing among middle school ado- lescent readers. Elementary School Journal, 120(3), 507–527. https:// doi.org/10.1086/707140

Ray, A., Graham, S., Houston, J., & Harris, K. R. (2016). Teachers’ use of writing to support students’ learning in middle school: A national survey in the United State. Reading and Writing, 29(5), 1039–1068. https://doi.org/10.1007/s11145-015-9602-z

Rietdijk, S., van Weijen, D., Janssen, T., van den Bergh, H., & Rijlaarsdam, G. (2018). Teaching writing in primary education: Classroom practice, time, teachers’ beliefs and skills. Journal of Educational Psychology, 110(5), 640–663. https://doi.org/10.1037/ edu0000237

Rocha, R., Filipe, M., Magalh~aes, S., Graham, S., & Limpo, T. (2019). Reasons to write in grade 6 and their association with writing qual- ity. Frontiers in Psychology, 10, 2157. https://doi.org/10.3389/fpsyg. 2019.02157

EDUCATIONAL PSYCHOLOGIST 71

Rogers, L., & Graham, S. (2008). A meta-analysis of single subject design writing intervention research. Journal of Educational Psychology, 100(4), 879–906. https://doi.org/10.1037/0022-0663.100.4. 879

Sackett, D., Rosenberg, W., Gray, J., Haynes, R., & Richardson, W. (1996). Evidence based medicine: What it is and what it isn’t. British Medical Journal, 312, 71–72. https://doi.org/10.1136/bmj.312. 7023.71

Saddler, B., & Graham, S. (2007). The relationship between writing knowledge and writing performance among more and less skilled writers. Reading & Writing Quarterly, 23(3), 231–247. https://doi. org/10.1080/10573560701277575

Santangelo, T., & Graham, S. (2016). A comprehensive meta-analysis of handwriting instruction. Educational Psychology Review, 28(2), 225–265. https://doi.org/10.1007/s10648-015-9335-1

Sawyer, R., Graham, S., & Harris, K. (1992). Direct teaching, strategy instruction, and strategy instruction with explicit self-regulation: Effects on learning disabled students¼ composition skills and self- efficacy. Journal of Educational Psychology, 84(3), 340–352. https:// doi.org/10.1037/0022-0663.84.3.340

Sfard, A. (2016). An interview with Anna Sfard. In A. Ovortrup, M. Wiberg, G. Christensen, & F. Hansbol, F. (Eds.), On the definition of learning (p. 323–330). University Press of Southern Denmark.

Shanahan, T. (2006). Relations among oral language, reading, and writ- ing development. In C. A. MacArthur, S. Graham, & J. Fitzgerald (Eds.), Handbook of writing research (pp. 171–183). Guilford Press.

Skar, G., Graham, S., & Huebner, A. (in press). The effects of COVID- 19 and online education on first grade students’ writing: A natural experiment. Journal of Educational Psychology.

Skar, G., Lei, P., & Graham, S. (in press). Handwriting fluency and the quality of primary grade students’ writing. Reading and Writing.

Snow, R. (1994). Abilities in academic tasks. In R. Sternberg & R. Wagner (Eds), Mind in context: Interactionist perspectives on human intelligence (pp. 3–37). Cambridge University Press.

Sperling, M. (1996). Revisiting the writing-speaking connection: Challenges for research on writing and writing instruction. Review of Educational Research, 66(1), 53–86. https://doi.org/10.3102/ 00346543066001053

Treiman, R. (Ed.). (1997). Spelling. Kluwer. Wijekumar, K., Beerwinkle, A., Harris, K. R., & Graham, S. (2019).

Etiology of teacher knowledge and instructional skills for literacy at the upper elementary grades. Annals of Dyslexia, 69(1), 5–20. https://doi.org/10.1007/s11881-018-00170-6

Wijekumar, K., Graham, S., Harris, K. R., Lei, P., Barkel, A., Aitken, A., Ray, A., & Houston, J. (2019). The roles of writing knowledge, motivation, strategic behaviors, and skills in predicting elementary students’ persuasive writing from source material. Reading and Writing, 32(6), 1431–1457. https://doi.org/10.1007/ s11145-018-9836-7

Wijekumar, K., Harris, K. R., Graham, S., & Meyer, B. (2017). We- Write: A teacher and technology supported persuasive writing tutor for upper elementary students. In S. Crossley & D. McNamara (Eds.), Adaptive educational technologies for literacy (pp. 184–203). Taylor & Francis.

Zimmerman, B., & Risemberg, R. (1997). Becoming a self-regulated writer: A social cognitive perspective. Contemporary Educational Psychology, 22(1), 73–101. https://doi.org/10.1006/ceps.1997.0919

72 S. GRAHAM

Copyright of Educational Psychologist is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.